pytorch/benchmark

Pull Requests Commits

Pin rotary-embedding-torch to 0.3.3

desertfire committed 2 years ago

df987fa4

Fix a100 bisection workflow (#2055)

xuzhao9 committed 2 years ago

9dea2810

Add Stateful/Stateless symbolic contexts, use fresh fake mode for dynamo backends

voznesenskym committed 2 years ago

83500ceb

Add the A100 bisection workflow (#2054)

xuzhao9 committed 2 years ago

8e5e3a0c

Add the test_bench userbenchmark (#2052)

xuzhao9 committed 2 years ago

7a8b39c8

Benchmark to save sample inputs to disk before running (#114163)

BowenBao committed 2 years ago

ec231246

Switch to check against expected result files for cpu inductor integration tests (#113668)

desertfire committed 2 years ago

ef244f0a

Switch to check against expected result files for dynamo_eager and aot_eager benchmark tests (#113559)

desertfire committed 2 years ago

ba4e45bb

Remove CI skip list for inductor integration tests (#113446)

desertfire committed 2 years ago

8e917699

remove GPT2ForSequenceClassification from ci skip list (#112100)

leslie-fang-intel committed 2 years ago

743294fa

Enable import following in MYPYNOFOLLOW (now MYPYINDUCTOR) (#113830)

int3 committed 2 years ago

1b2034ee

Fix bench w/ iobinding; Remove cpu fallback (#113703)

BowenBao committed 2 years ago

10d61e8e

Minor comment cleanup, typos (#113898)

janeyx99 committed 2 years ago

6da25201

Make _inductor/fx_utils.py, _dynamo/utils.py pass follow_imports typechecking (#113722)

int3 committed 2 years ago

4f134a40

Skip the bisection when the start and end hash are the same (#2044)

xuzhao9 committed 2 years ago

42785452

re-add option to log all guard check fails (#113585)

williamwen42 committed 2 years ago

ff42ab82

Drop dynamo-specific type hints on Tensor in favor of type-ignores (#113720)

int3 committed 2 years ago

361a7756

Register grid buffers for Super_SloMo. (#2038)

ysiraichi committed 2 years ago

fa41cb26

Fix stable_diffusion_xl fp16 (#2032)

BowenBao committed 2 years ago

c7cb5aa1

Remove useless lambdas (#113602)

Skylion007 committed 2 years ago

876017cb

ruff apply rule PLW1510 to find silent subprocess errors (#113644)

Skylion007 committed 2 years ago

10693103

Add 01-ai/Yi-6B (#2037)

MaanavD committed 2 years ago

d1d71d51

Log both torch.version.git_version and torch.__version__ (#2039)

xuzhao9 committed 2 years ago

7d3f6f0a

Catch OSError when calling inspect.getfile (#113671)

yanboliang committed 2 years ago

10b18161

replace uses of numpy.long with int

igorsugak committed 2 years ago

a1c4fdf3

Make {testing,debug_utils,utils}.py pass follow_imports typechecking (#113519)

int3 committed 2 years ago

317b36a5

TorchMultimodal CLIP benchmark example (#2033)

ebsmothers committed 2 years ago

4d01954d

Force fp16 for vision_maskrcnn inference (#113110)

eellison committed 2 years ago

f7a62d6f

Update the benchmarking script to clone an eager model (#113046)

desertfire committed 2 years ago

27968712

Enable typechecking for utils.py (#112971)

int3 committed 2 years ago

15ce7cd8

Older